by
Y.S. Demin
Institute of Developmental Biology
Academy of Science
Moscow, U.S.S.R.
The term heterosis (or hybrid vigour) characterizes the increased ability or a hybrid as compared to the parental forms. The term ability means the favourable changes in hybrid characters when compared to the abilities of P (Parental) in one or several characters. This phenomenon was discovered in the 18th century by Veldeiter during experiments on tobacco hybridization and later confirmed by many selectionists working on hybridization of plants and animals. It was noticed that the main value of inbreeding - retention of desired characters - is connected to an equal degree with the risk of their diminution. Darwin (1876) was the first to attempt to explain the theoretical fundamentals of heterosis. He concluded that cross-pollination usually produced a favourable effect, and spontaneous pollination a detrimental one. Darwin's works on comparison of spontaneous pollination and cross-pollinated plants gave rise to many investigations on selection. By the time of the re-discovery of Mendel's laws, considerable practical material was collected which confirmed the conclusion made by Darwin and also the two very important earlier observations made by Kolreuter (1763,1766) (hybrid ability is connected with the degree of genetic difference between their parents; hybrid vigour is of special importance in the process of evolution).
At present the phenomenon of hybrid ability has been investigated adequately only from the point of view of application, and no general theory of genetic heterosis has been created. From a number of concepts, two theories have been most successfully developed: dominant linkage factors and super-dominance. There are experimental data in favour of both concepts. However, there are facts which do not completely conform to either of them. Eventually, when a general theory of heterosis is developed, each concept will have its proper place.
This concept proceeds from the assumption that hybrid power develops as a result of action and interaction of allelic and non-allelic dominance and very often, linked genes.
According to the theory, heterozygosis proper is not an obligatory factor for heterosis. Theoretically, the specimens, being homozygous due to certain favourable factors, must be equal in power to the specimens which are heterozygous due to the same factors. Let us consider a hypothetical model of heterosis conditioned by dominance:
P AoAoB1B1CoCo 1 1 EoEo × A1A1BoBoC1C1ooE1E1
F Ao A1BoB1CoC1 o 1EoE1
Let us assume that the alleles with index 1 are dominant and that each recessive genotype (XoXo) introduces one unit into a given character and each dominant (X1X1 or X1Xo) introduces two units. Then the contribution of units of the given character in each parental genotype will be respectively:
P(1 + 2 + 1 + 2 + 1) × (2 + 1 + 2 + 1 + 2)
and in F1 (2 + 2 + 2 + 2 + 2) or
PF1 7 × 8
↓
10
If a theoretical specimen, homozygous by all the dominant alleles, is assumed to exist, this specimen will have the same number of units of the given character as hybrid F1 :
A1A1B1B1C1C1 1 1E1E1 = 2 + 2 + 2 + 2 + 2 = 10; i.e.,
according to this series, heterozygosis is not obligatory for heterosis.
One objection to the theory of dominant factors is that it is practically impossible to produce pure lines with abilities equal to hybrids. However, mathematical calculations (similar to those in polyhybrid crossing) indicate that segregation in F2, homozygous by the given factors, is equal to (¼)n, where “n” is the number of loci controlling the given character. But, in corn, for example, 30 genes are known which are responsible for yield. Therefore, of 430 plants of F2 only one plant will be homozygous. The area of the earth is insufficient to produce this number of plants.
Another important objection is the obviously symmetrical distribution of the given character in F2 received from F1. In other words, the distribution of values of quantitative characters indicating heterosis is symmetrical, which is similar to to the distribution in which dominance is not observed. However, according to the theory of dominance, the distribution must deviate from the normal, the asymmetry being considerable. Later it was proved that lack of asymmetry may be caused by the influence of the environment and other factors (incomplete dominance, increase of gene number, etc.).
Jones (1918), who was the first to formulate a theory of heterosis on the basis of the phenomenon of dominance, attached great importance to linkage as an additional factor which caused heterosis. According to him, linkage can present a factor hindering development of chromosomes with many dominant loci (and consequently hindering segregation of homozygotes) and reducing the expected asymmetry of splitting populations.
It must be pointed out (it follows from the above) that so far we have discussed interaction of allelic factors; and now we shall consider the interpretation of heterosis, which was the basis of the dominance hypothesis. What is meant here is the idea of screening detrimental recessives by dominant alleles. Although this idea is correct from the point of view of viability and survival, now it has been proved that it cannot be considered the main cause of heterosis. It is known that a theoretical value of heterosis for corn calculated on the basis of screening detrimental recessives by dominant alleles should not exceed 5 percent, whereas the best interline hybrids exceed the original grades by 30 percent in yield. Thus a certain genetic cause can explain adverse consequences of inbreeding and restoration of normal viability in crossing inbred lines; but not heterosis as a phenomenon.
The consideration of nonallelic interaction of genes is an important problem of the heterosis theory based on dominance. One of the main types of interaction is a complementary interaction of genes in heterosis. If both dominant alleles A1 and B1 1 are necessary for synthesis of any metabolite, then homozygotes AoAoB1B1 and A1A1BoBo are not able to synthesize this product. The ability to synthesize this metabolite, which is inherent in the hybrids of these forms, may be considered as a form of heterosis. The hybrid between two kinds of tomatoes (Current and Johnannesfeier) is an example of this type of interaction. One variety had a reduced ability for synthesis of pyridoxine (vitamin B6) and the other for the synthesis of nicotineamid resulting in retarded growth of roots. The hybrid of these two varieties obtained the ability to synthesize both vitamins and, therefore, its rootstock developed more intensively. Thus, in this case, heterosis is the effect of the presence of active alleles in the heterozygote originating from different loci.
Now, it can be seen that the concept of dominant factors consists of at least three points: an additive effect of linked dominant factors; a suppressing action of dominant factors in respect to detrimental recessives; and non-allelic interaction of genes. The notion of dominance is somewhat broader here, because it is not applied to interaction of allelic pairs, but to interaction of parental genotypes, affecting the development of any character in the hybrids. However, in the final analysis, this development is determined by dominant genes.
The basis of the theory of superdominance is the idea that in some cases interaction of two allelic genes may result in heterozygotes AoA1 surpassing both homozygotes in power: AoAo and A1A1. According to this series, in the case of such interaction of alleles, it is not possible to produce a homozygote which could be equal in power to a heterozygote characterized by superdominance. In other words, heterozygosis is necessary for complete development of heterosis. This does not mean, however, that the dominant linkage factors of growth do not contribute to development of heterosis; according to the theory of superdominance, they alone cannot be the cause of a complete development of heterosis.
The superdominance theory assumes that both alleles have somewhat different functions in the heterozygote, and the effect of their interaction is based on this phenomenon. The theory is based on facts which show the superdominance of qualitative characters, conditioned by heterozygosity by one locus (single gene heterosis). There are four types of allelic interaction which may result in one-gene heterosis. It is quite possible that these types of interactions may occur in various combinations in the same hybrid.
In co-dominance of both alleles, the heterozygote receives the products of activity of both genes. It may be represented as follows:
P AoAo→Xo; A1A1→X1; where X = products of gene.
I AoA1→ Xo + X1
This type of one-gene heterosis is seen in many types of plants. For flax, in particular, it has been proved that the alleles of immunity for rust are codominant, and therefore the presence of two alleles in a genotype makes F1 much more immune to disease than homozygous forms. The phenomenon of one-gene heterosis has been found in carp. Inheritance of blood groups in human beings and animals conforms with this phenomenon. For chickens and hogs, heterosis is indicated in case of heterozygosity by certain blood groups. However, in this case heterozygosis seems to be conditioned by the effect of loci closely linked with the genes of blood groups. A pleiotropic effect of genes determining the blood groups may be considered as the second explanation.
Alternative ways of synthesis, which is the basis of more extensive reaction norms in heterozygotes as compared to homozygotes, differ from the previous case, mainly by the role of environmental conditions in the development of allelic effect. For example, assume that allele R1 produces pigment, maximum productivity being at 27°C, and allele R2 produces the same pigment in optimum quantity at 10°C. Due to absence of dominance, the organism, heterozygous by these alleles, will be better adapted to variations in the environmental conditions than both homozygotes. Experiments have proved that in some cases hybrids react weaker than inbred lines to variations of environment. The concept of optimum quantity is based on the assumption that a homozygote by one allele produces too little of a necessary metabolite, whereas it produces too much of it by the other allele. The heterozygote will have an intermediate ability to synthesize and produce the optimum quantity. This type of heterosis, determined by one gene, is widespread. The type of heterosis effect is found in Drosophila, barley, corn, etc. What is meant is a gene dosage, namely, quantity of products developed by a gene. But in connection with the fact that this gene dosage depends to a great extent on a genotype environment; i.e. on pleiotropic effect of other genes and on gene modifiers. It is difficult to present a final experimental proof for heterosis of this type.
The hypothesis of hybrid substance states that a homozygote by one allele gives product Xo (AoAo→Xo); a homozygote by the other allele, A1(A1A1→X1); and a heterozygote develops a hybrid product (AoA1→Xo+1). In some cases, the heterozygote may have all three products: XoX1 and Xo+1. A great number of hybrid material of this type has been found in heterozygotes (three types of haemoglobin, one of them being hybrid, are found in sunfish; hybrid haemoglobins and enzymes are found in birds); however, their connection with heterosis by a single gene has not been proved.
Another example in favour of the superdominance hypothesis is the double interlinear hybrid of corn. Such hybrids are produced by means of combination of four non-related lines. If the heterosis of common interlinear hybrids (being parent forms of double hybrids) is the result of suppression of the effect of deterimental recessives by dominant alleles in crossing common hybrids, a considerable part of undesirable recessives must transfer into homozygous state; i.e. double hybrids must always yield to common hybrids in power. In practice, double hybrids of corn do not yield power to the best of common hybrids, which can be easily explained by the theory of superdominance.
Heterosis is superiority of a hybrid over the parent forms in the rate of development of one or more characters. The extent of development of a character is the result of development, characterizing this process quantitatively. Deviations in the extent of development of a character in hybrids may result in increase or decrease in manifestation, which are termed positive or negative heterosis. They supplement each other, providing an explanation to some effects of hybrid vigour. Both theories proceed from a common initial basis, namely, hybrid vigour which is the result of gene interactions. Interactions may be allelic and non-allelic. This result may be a consequence of interaction of separate genes as well as of gene groups. The heterosis effect of gene groups has been studied in populations of Drosophila. As a result of some structural modifications in chromosomes, such complexes of genes may appear which are not separated by crossing-over. These groups supplement each other when joining in a heterozygote, providing for better ability of heterozygotes as compared to the corresponding homozygotes. It is to be concluded that natural selection in a population develops in such a way that gametes joining in pollination supplement each other in the best way and produce a generation of stronger constitution.
Heterosis in populations, conditioned by reciprocity of alleles and gene complexes and providing for an optimum constitution, is called euheterosis (Dobzhansky). In practice man very often encounters excessive heterosis; i.e. development of a character to such an extent that it may be utilized only for economic or experimental purposes, but is not important in natural conditions. It is clear that both types of heterosis are but different manifestations of the same phenomenon.
The role of cytoplasm is of significance in some cases of heterosis. As is known, the character of a genotype changes during development, depending on the properties of zygote cytoplasm. The role of cytoplasm in heterosis can be explained as follows: on reciprocal crossing of two lines A × B and B × A, heterosis is revealed by similar properties in the hybrids of one of the crossings and is not revealed in the hybrids of the other.
The theory of genetic balance formulated by Turbin (1966) attempts to coordinate the ideas available into a general theory of heterosis. The theory of genetic balance proceeds from the concept that if the development of a character is the result of genetic balance (which is the equilibrium attained between oppositely directed actions of different genes on the given character), the removal, modification or substitution of some of them will necessarily give advantage to the factors of opposite action. In organisms with broken genetic balance it may cause change in the extent of development of some other characters. Proceeding from this concept, heterosis may be developed as one of the consequences of a modified genetic balance in hybrids produced from non-related lines. It is necessary to note that sterility of genetically non-balanced forms in some cases does not exclude their practical application (only F1 is used).
As can be seen, the theory of genetic balance provides a too general explanation of the causes of heterosis; but it does not specify the role of some other types of interaction of heredity factors which determine the phenomenon, and which are the components of genetic balance of hybrids. This theory is as correct as any other to explain the general notions, but at present it is hardly plausible as a general theory of heterosis.
The phenomenon of heterosis has been observed in all the species studied. In many cases, heterosis is so obvious that there is no need to resort to statistical analysis to demonstrate its value. This is particularly true of corn hybrids.
When two hybrid lines of corn are crossed, very often hybrids F1 produce twice as much seed yield as the parents. The use of hybrid seed is now the main method of growing corn for grain and for silage. In order to produce hybrid seed, inbred 1 lines of good varieties are obtained which meet the requirements of the given climatic region (an inbred line is created for 5–6 years by means of random pollination). In selection of lines, their qualities are estimated in connection with the properties which should be obtained in a future hybrid organism. Inbreeding cannot be effective if not accompanied by selection. Having created a great number of lines, crossing is begun. Interlinear hybrids of the first generation are estimated by a heterosis effect; proceeding from this, the lines of better combining ability 2 are selected and then reproduced at a greater rate for production of hybrid seed. At selection stations, work on production of inbred lines and estimation of their combining ability is carried out continuously. The more valuable lines are created as soon as it is possible to select better hybrid matching with necessary combination of characters. It has already been said that at present double hybrids of corn are used. The process of matching common hybrids for the purpose of obtaining the most productive double hybrids is a very important stage in the process of selection. The best results are achieved in crossing the lines originating from different varieties or strains. Corn is an excellent example to show that, for the successful development of hybrid seed growing, it is first necessary to investigate how long inbreeding should be carried on in order to achieve homozygosity by the group of genes which are of interest to us; and second, to develop methods for rapid estimation of their combining ability.
The above facts concerning corn are likely to be true; to a certain extent, in obtaining hybrids of other cross-pollinating plants and animals as well. Crossing of inbred lines from one or various breeds has been widely used now in the field of poultry and pig breeding. It is necessary to note that wide scale utilization of hybrids in cattle breeding is possible only at the highest level of pure-strain stock farming where there is an availability of valuable breeds.
It is clear that in most cases the inbred lines will always have lower indexes than the strains. Heterosis is evident only when the interlinear hybrid exceeds not only its parent lines but also the varieties or breeds from which these lines generate.
6.1.1 Combining ability
In crossing aimed at obtaining heterosis effect particular attention should be paid to combining capacity, distinguishing between general and specific combining ability. The first is characterized by an average amount of heterosis, observed in all hybrid combinations; the second, by deviation from this amount in one or another separate combination. For determination of general and specific combining ability, the form under test and the corresponding analyses (tester) are crossed. Moreover, for determination of general combining ability it is better to use analyses with a wide genetic basis (random pollination grade or corresponding animal population). The specific combining ability of the form being tested is estimated in relation to any form with which it is to be crossed later. This second form is an analysis. Choice of analyses depends on the purpose of the lines tested, i.e. whether they will be used to substitute the line in the existing hybrid combinations, or for the production of new hybrids.
The analysis for determination of combining ability of the material being selected is used for a system of crossings. For this purpose various systems of crossing are used, which are in fact various methods of determination of combining ability.
In plant growing, the following four methods are used: diallelic crossing, topcross, poly-cross, and random pollination. Of these methods only top-cross can be used in cattle breeding.
It is evident that the combining ability (general and specific) can be improved as a result of selection of recombinant forms having new combinations of genes. In this way it is possible to obtain new genotypes of a higher combining ability. Hybridization and selection are the means with which favourable complexes of hereditary factors are concentrated in a population. Effectiveness of the above means, in practice, that selection depends to a great extent on the methods applied for detection of genotypes which differ in combining ability. The complete programme of selection for determination of combining ability, the purpose of which is to obtain new components for crossing or for improving the existing lines, must provide for certain alternations of ways to produce new genetic combinations, their breeding estimation for combining ability, and selections of the best genotypes.
Depending on the direction of selective breeding of new lines, improvement of existing lines, or selection for determination of general or specific combining capacity, there are various selection programmes in existence. Various kinds of periodic selection are used, as well as convergent improvement, cumulative selection and gamete selection.
6.1.2 Periodic selection
Periodic selection provides for alternation in crossing of the material selected with the inbred material and selection in the intermediate period. The selection is carried out on the basis of estimation of species selected by the characters which are of interest.
6.1.3 Convergent improvement
This is a method suggested for corn selection. The theoretical basis of the method involves the fact that heterosis is considered as the result of a favourable effect of dominant linkage factors. By this method, using the back-cross of common hybrid A × B for the parental lines, new self-pollinated lines A' and B' are obtained, each of them carrying a number of genes of the second parental line. In this way the accumulation of favourable dominant genes in new parental lines A' and B' is ensured.
6.1.4 Cumulative selection
Cumulative selection is also based on the theory of dominance. This method is used to accumulate favourable dominant genes by crossing the lines which have excellent combining ability, and selecting the best combinations.
6.1.5 Gamete selection
Theoretically this is related to the phenomenon of superdominance. This method provides for improvement of the selected self-pollinated line on the basis of recombination, achieved by random selection of gametes of the strain. In order to obtain a new improved line, the line selected is crossed with a definite strain or hybrid, assuming that the offspring will differ from the initial line only by such properties (or gametes) which were introduced by the strains.
The problem of hybrid vigour retention, i.e. constant development of heterosis is a sequence of generations, has several theoretical approaches, some of which are applied in practice in plant growing. However, in animal breeding, the problem of retention of hybrid vigour remains to a considerable degree unsolved.
The theoretical basis of heterosis retention is accumulation in a population of balanced, closely-linked gene blocks, the combination of which in the heterozygote leads to heterosis. In other words, it is necessary to avoid segregation of the valuable genotype obtained as a result of crossing.
One of the approved methods of retention of heterosis is asexual reproduction of hybrids. This method of retention of hybrid vigour is widely used in selection of fruit plants.
Another form of asexual reproduction in plants is apomixis. In some cases of apomixis the seeds develop from diploid cells, both stages of meiosis being omitted, and the generation of these seeds is genotypically identical to the mother plant. Retention of heterosis through apomixis is used in growing blackberry, roses, meadow grass, citrus and many other plants.
Retention of hereditary heterosis may occur as a result of combination of lethal genes and certain chromosomic arrangements. In a specimen, heterozygous by inversion, the crossing-over inside of an inverted section leads to development of sterile gametes. As the crossovers obtained inside of an inverted section are suppressed due to sterility of crossover gametes, the heterozygosis of the inverted section becomes permanent. This type of balanced lethality is observed in hymenopteran insects.
Besides, heterozygosis can be retained or improved by means of a definite combination of recessive lethal genes with other types of chromosomic arrangements - translocations.
Retention of hybrid vigour sometimes is obtained with the aid of polyploidy. The practice of plant culture shows that increased vigour of interspecific hybrids is observed, not only in diploid hybrids which can be sterile, but also in amphidiploid forms, in which the chromosomic combination is doubled. Interspecific hybrids with a double combination of chromosomes are very often fertile, and their descendants continue manifesting heterosis developed as a result of the genotypes from two species.
Especially of interest in fisheries is a very promising method of retention of heterosis with the help of gynogenesis. 1 It has been proved that in some cases the gynogenetic development of the zygote in fishes results in the emergence of clones, i.e. forms genotypically absolutely identical to the mother organism. In other words, meiotic segregation is absent in this case. Thus, development of gynogenesis in fishes on a commercial scale might facilitate the solution of the problem of retention of heterosis.
Heterosis is hereditary development of favourable effects resulting from hybridization. Briefly, it may be said that the power lost in the process of inbreeding tends to be restored in crossing. Diminution of power in inbreeding is usually connected with increase of homozygosity by detrimental genes. Hybrid power, on the contrary, is not connected directly with increase of heterozygosity.
Thus, there are two points of view on genetic causes of heterosis, each supplementing the other. One of them - the theory of favourable dominant factors - connects heterosis with allelic and non-allelic interaction of genes. The other - the theory of superdominance (or monofactorial heterosis) - is based on co-dominance and interaction of alleles.
The scheme of the development of heterosis is to obtain a combination of lines which have the desired properties. The combining value of this line is estimated by means of crossing it with another line of an adequately heterozygous population. The hybrids F1 produced by one interlinear crossings very often cross between themselves, etc. Selection of forms for crossings with the purpose of obtaining hybrid power is still based mainly on empirical matching of pairs. Development of a general theory of heterosis could help in planning the experiments on crossing in such a way that the effect desired could be foreseen. Development of a general theory of heterosis is retarded by the fact that we do not know all the details of gene interaction.
The application of heterosis in fisheries is rather promising. High fecundity in fishes allows a considerable increase of yield to obtain adequate lines for crossing. A theoretical possibility of retention of heterosis in fishes with the aid of gynogenesis opens wide perspectives of commercial heterosis in fisheries. However, we must stress that wise use of hybrids in fisheries (as well as in animal breeding) in general) is possible only at high levels of pedigreed stock farming and with valuable breeds.
For fisheries, the combination of hybrid power and sterility can be of potential value. Tropical Tilapia can be taken as an example. This is a fresh-water fish, and it reproduces so rapidly that over-population is often observed in ponds. In an over-populated pond, a fish gains about 230 g per year. However, some interlinear hybrids develop powerful, rapidly growing, sterile males. In the ponds populated with such hybrids there is no over-population; the speed of growth, determined by heterosis, reaches its complete development and the annual gain is 1360 g.
In conclusion we may say that although the phenomenon of hybrid vigour has been well investigated as to its application, the theoretical aspect of the problem is full of difficult genetic questions. We think that insufficient knowledge about genetic causes of heterosis is connected with poor “resolution ability” of the objects used in agricultural practice for experiments (Drosophila). Probably some success in this respect could be achieved in studying heterosis in such subjects as fungi, which have short life cycles, are adequately studied genetically, and allow an analysis of many hybrid descendants at one time.
Darwin, C., 1876 Cross and self fertilization in the vegetable kingdom. London.
Dobzhansky, T., 1950 Genetics of natural populations. XIX. Origin of heterosis through natural selection in populations of Drosophila pseudoobscura. Genetics, 35: 288–302
Jones, D.F., 1918 The effects of inbreeding and crossbreeding upon development. Bull.Conn.agric. Exp. Sta., (207): 100 p.
Kirpichnikov, V.S., 1967 The general theory of heterosis. Genetika, (10): 166–80
Kolreuter, J.G., 1763 Vorlaufige Nachricht von einegen das Geschlecht der Pflanzen betreffenden Versuchen und Beobachten, Fortsetzungen 1. (Ostwald Klassiker der exacten Wissenschaften, 41, Leipzig, 1893)
Kolreuter, J.G., 1766 Vorlaufige Nachricht von einegen das Geschlecht der Pflanzen. Leipzig.
